POTASSIUM AND AMINO ACID TRANSPORT IN HUMAN LEUKOCYTES EXPOSED TO PHAGOCYTIC STIMULI

Published October 1, 1974

P O T A S S I U M AND AMINO ACID T R A N S P O R T IN H U M A N LEUKOCYTES EXPOSED TO PHAGOCYTIC S T I M U L I

PHILIP B. DUNHAM, IRA M. GOLDSTEIN, and GERALD WEISSMANN From the Department of Medicine, New York UniversitySchool of Medicine, New York 10016. Dr. Dunham's permanent address is Department of Biology, Syracuse University, Syracuse, New York 13210.

ABSTRACT

During the process of particle ingestion, polymorphonuclear leukocytes (PMNs) internalize large portions of their surface membrane, but the relationship between internalization and carrierTHE JOURNAL OF CELL BIOLOGY . VOLUME

63, 1974

mediated membrane transport remains unclear. Tsan and Berlin (1971) demonstrated that although the phagocytosis of polyvinyl toluene latex particles by rabbit PMNs and alveolar macro-

• pages 215-226

215

Downloaded from jcb.rupress.org on July 31, 2015

Influxes of potassium and amino acids were measured in suspensions of human polymorphonuclear leukocytes (PMNs) under resting conditions and after various phagocytic stimuli. Both ouabain-sensitive (or pump) and ouabain-insensitive (or leak) influxes of K were determined. In 5 m M external K, mean total K influx was 0.69 nmol/10 e cells × min, of which 52% was ouabain-sensitive. Ouabain binding was irreversible, and, as in erythrocytes, was inhibited by K. At external concentrations of 0.1 raM, influxes of lysine and leucine were entirely carriermediated, with means of 0.021 nmol/l08 cells x min, and 0.019 n m o l / l 0 s cells × min, respectively. After incubation of P M N s with zymosan or latex particles, the K pump was reduced more than 60%, whereas amino acid influxes were inhibited only by 30%. P M N s were also exposed to cytochalasin B before challenge by particles: the drug prevented phagocytosis but not surface binding of zymosan, nor did it influence transport of K or amino acids. After pretreatment of P M N s with cytochalasin B, interaction of zymosan with their surface resulted in the same degree of inhibition of influxes of K and amino acids as when the cells were permitted to phagocytose the particles. In contrast, exposure of P M N to latex particles, which do not bind to cytochalasin B-treated cells, after pretreatment of cells with cytochalasin B did not result in inhibition of influxes. Treatment of cells with colchicine had no effect on either membrane transport or its inhibition after exposure to various phagocytic stimuli. These results indicate that the surface membranes of P M N s are functionally heterogeneous with respect to the association of transport sites for the different solutes. Moreover, loss of "specific membrane functions from phagocytosing cells may result from the surface-attachment phase of particle-cell interactions, since the interactions of zymosan particles with P M N s in the absence of phagocytosis also inhibited transport of solutes.

Published October 1, 1974

MATERIALS AND METHODS

Preparation of Leukocytes Leukocyte suspensions were prepared as previously described from venous blood obtained from healthy young male donors (Weissmann et al., 1971). Blood was drawn into heparinized syringes and mixed with 0.25 vol of a 6% dextran solution (mol wt 190,000-230,000; Pharmachem, Bethlehem, Pa.) in 145 mM NaCI. Sedimentation of the red cells (facilitated by the dextran) was

216

allowed to proceed for 30 min; the leukocyte-rich supernate was removed and centrifuged at 150 g for 7 rain at room temperature. The pellet was suspended in 145 mM NaCI to which 3 vol of distilled water were added to lyse the remaining red ceils. After mixing for 20 s, the suspension was restored to isotonicity by addition of an appropriate volume of 615 mM NaCI. The leukocytes were sedimented by centrifugation at 150 g for 7 rain, resuspended, and washed by an additional centrifugation in phosphate-buffered saline containing calcium and magnesium (NaC1 137 mM, KCI 2.7 mM, CaCI2 0.6 mM, MgCI~ 1.0 mM, Na2HPO4 8.1 mM, KH~PO, 1.5 mM, pH 7.4). The final leukocyte suspensions were more than 85% PMNs, with a platelet-to-leukocyte ratio of approximately 5: 1. Suspensions containing 98-99% PMNs were prepared similarly after initial fractionation of venous blood on Hypaque/Ficoll gradients, according to the method of B~yum (1967).

Experimental Procedures." Preincubation Leukocyte counts were performed and suspensions were adjusted to contain 2 4 × 106 cells/ml. Cells were dispensed in 0.5-ml volumes into disposable plastic tubes. Autologous serum was added to a final concentration of 10%. The cells were incubated with various pharmacologic agents and particles at 37°C in a total volume of 1 ml. Cells prepared in this manner were shown previously to be over 90% viable and capable of active phagocytosis (Weissmann et al., 1971).

Potassium Influx Unidirectional influxes of K were measured with 4~K. The *~K medium was prepared in buffered saline identical in composition to that given above except for the absence of K. Autologous serum was added to a final concentration of 10%. 4~KC1 (New England Nuclear, Boston, Mass.) or '2K2COa (ICN Corp., Chemical & Radioisotopes Div., Irvine, Calif.) was added to give a final 42K concentration of 5 mM. (For K2CO3, the pH was restored to 7.4 with HCI). In some experiments, the K medium contained various pharmacologic agents, as described in Results. In experiments designed to measure influx of K and amino acids simultaneously, medium also contained [14C]amino acids, as described below. At the end of the preincubation, cells were sedimented by centrifugation (7 rain, 150 g) and the supernate was removed by aspiration. The cells were resuspended with Pasteur pipettes in 1 ml of the buffered saline (at 37°C) containing 42K. This resuspension procedure also served to separate clumps of cells which may have formed during the preincubation. The tubes were incubated at 37°C, generally for 10 min. The '2K influx was stopped by addition of 2 ml of cold buffered saline and immersion of the tubes in an ice bath. The cell suspensions were centrifuged i n the cold (4°C) at 300 g for 10 rain and the supernate removed by aspiration. The cells were washed

THE JOURNAL OF CELL BIOLOGY - VOLUME 63, 1974

Downloaded from jcb.rupress.org on July 31, 2015

phages resulted in internalization of an estimated one-third to one-half of the surface membrane, the rate of carrier-mediated influx of several organic solutes (lysine and adenosine in macrophages; adenosine and adenine in P M N s ) was not appreciably reduced. It was suggested that transport rates remained intact not because internalized transport sites were replaced, but rather because transport sites on the surface m e m b r a n e were spatially separated from areas of the m e m b r a n e that had become internalized during phagocytosis. More recently, U k e n a and Berlin (1972) showed that transport rates of adenine and lysine were reduced 40% after phagocytosis of latex particles by rabbit P M N s pretreated with colchicine or vinblastine (which interfere with microtubule function) (Weisenberg et al., 1968; Berry and Shelanski, 1972); whereas transport rates of untreated cells were unaffected. Colchicine had only a small effect on solute transport by resting (nonphagocytic) cells. The authors concluded that the integrity of microtubules was required for the apparent topographic separation of areas of the m e m b r a n e involved in transport and phagocytosis, or at least for the cells' ability to achieve functional separation. We have reexamined the relation between transport and phagocytic functions in human P M N s by measuring transport of potassium and two amino acids in resting cells and in cells after ingestion of two different kinds of particles. Our results indicate that there is a complex, functional relationship between transport sites, particle-binding sites, and area of the surface m e m b r a n e involved in phagocytosis. Particle ingestion resulted in a large inhibition of K transport; amino acid transport was also inhibited but to a lesser extent. In cells treated with cytochalasin B, which inhibits phagocytosis, surface binding of particles had the same effect on transport as internalization of surface membrane. Finally, colchicine was without effect on these aspects of m e m b r a n e function.

Published October 1, 1974

cpm/liter of the medium, and (K)o = K concentration of the medium in nmol/liter. The cells were assumed to be steady-state with respect to K concentration during the 10-min flux period. In general, this appears to be a good assumption, though some of the pharmacological agents and other treatments m a y cause slight losses of K.

Amino Acid Influx Unidirectional influxes of amino acids were measured like the K influxes. When K and amino acid influxes were measured simultaneously, a [14C]amino acid, lysine or leucine (New England Nuclear), usually at a final concentration of 0.1 raM, was added to the medium containing 42K. The cells were incubated for 10 min at 37°C and the flux was stopped by addition of ice-cold buffered saline, as described above. The cells were centrifuged and washed three times at 4°C. They were then extracted in 2% Triton X-100 and 42K radioactivity was determined from Cerenkov scattering. Only the emission from 42K was detected, since the beta emission of x4C is not detected in the absence of fluor. After 1 wk (more than 10 half-lives of 42K), 10 ml of Bray's solution (Bray, 1960) was added to the counting vials containing the cell extracts and dilutions of the medium, and the radioactivity of the [~4C]amino acid was determined. Each sample was counted to at least 10,000 counts. With the use of an external standard it was determined that there were negligible differences in counting efficiency between samples. Unidirectional influxes of amino acids were calculated exactly like the K influxes.

Phagocytosis i~K = R t . (K)o Ro

(I)

where: i ~ = unidirectional K influx in nmol/106 cells z min, R, - cpm of cell extract/10 s cells x min, Ro -

TABLE I

Unidirectional K Influxes in Human PMNs Related to Numbers o f Cells and Times o f Exposure to 42K Number of cells

Exposure time

K influx

x I0s

min

nmol/min x 106 cells

2 2 4 4 8 8

10 30 10 30 10 30

0.57 0.44 0.57 0.53 0.60 0.53

Fluxes are means of duplicate determinations.

The following particles were employed: zymosan (ICN Nutritional Biochemicals Div., International Chemical G Nuclear Corp., Cleveland, Ohio) and polystyrene latex, 1.10 /~m in diameter (Dow Chemical USA, Midland, Mich.). Cells were exposed to particles for 30 min at the following particle-to-cell ratios: zymosan, 10: 1; latex, 100: I.

Statistical Test Levels of significance of differences between relative fluxes after various treatments were determined with the randomization test for matched pairs (two-tailed), a nonparametric test with 100% power-efficiency (Siegel, 1956). Differences at the 0.06 level were regarded as significant.

Compounds Ouabain and colchicine were purchased from Sigma Chemical Co., St. Louis, Mo. Cytochalasin B was obtained from Imperial Chemical Industries, Cheshire, England, and dissolved at a concentration of 50 g g / m l in phosphate-buffered saline containing 0.1% dimethylsulf-

DUNHAM, GOLDSTEIN, AND WEISSMANN Transport of K and Amino Acids in Leukocytes

217

Downloaded from jcb.rupress.org on July 31, 2015

free of extracellular radioactivity by resuspension in 2 ml of ice-cold buffered saline and centrifugation for 10 min. This procedure was repeated two more times for a total of three washes in cold saline. After three washes, the radioactivity of the supernate was negligible. As will be shown below, K and amino acid fluxes are reduced by 90% in the cold. Therefore, the washing procedure removes mainly extracellular isotope, and the radioactivity of the washed pellets of cells gives a good approximation of the influx during the incubation at 37°C. The washed pellets of cells were extracted in 0.5 ml of 2% Triton X-100. The extract was transferred quantitatively to a glass counting vial (27 × 58 m m ) by rinsing the tubes with additional portions of Triton X-100 solution. The radioactivity of the 42K was determined from the Cerenkov scattering in a Beckman LS-100 liquid scintillation counter (Beckman Instruments, Inc., Spinco Div., Palo Alto, Calif.). Total counting time never exceeded 45 min, so correction was not made for radioactive decay. Radioactivity of the flux medium was determined at the same time on a volume of an appropriate dilution equal to the volume of the cell extracts. The influx of 42K was a linear function of the number of cells per milliliter (between 2 x 106 and 8 x 106), and nearly a linear function of time (between 10 and 30 min), as shown in Table i. Therefore in 10 min no correction was made for back flux of "2K, and this time was chosen as the standard time for measurement of 4ZK influx. The unidirectional influx of K was calculated from radioactivity in counts per minutes of the medium, K concentration of the medium, the radioactivity of the cells, and the number of cells, by the equation:

Published October 1, 1974

oxide (Matheson Coleman & Bell, East Rutherford, N. J.). RESULTS

Potassium I n f l u x Table I demonstrates that, under these conditions, K influx is a linear function of cell number up to 8 × 106 cells in 1 ml of suspension medium, and is approximately a linear function of time up to 30 rain. Fig. 1 shows the time course of 42K influx, expressed as nanomoles of intracellular K exchanged per l0 s cells. If the exchange is a first-order process involving a single intracellular compartment, the following relation should be obeyed:

R e = R~(1 - e -kt)

or R__~)=_kt

(2)

where Rt is radioactivity, or amount of K exchanged, in the cells at time t, R® is the radioactivity at infinite time or the total amount of exchangeable K, and k is the rate constant of the net influx of 4ZK (equivalent to the sum of the rate constants for the unidirectional inward and outward fluxes of 4ZK). 40

30

1.0

w

0.7

~

0.5

~ zo -~ 0 . 2 tu

,

I0

01/ • 0

I I0

i

I 30

i

I

I 60

I

i IO

I 50

J

I 60 MINUTES I 90

I 90

I

i 120

I

i I 20

MINUTES

FIGURE I Time course of 42K influx in human PMNs, expressed as nanomoles of K exchanged per 106 cells. Inset shows data recalculated as 1 - (Rt/R®) where Rt is the radioactivity of the cells at times t and R= is radioactivity at time infinity, assuming that a total of 45 nmol/10 s cells of intracellular K is available for exchange with 4~K. See text for further explanation. Points represent means of two determinations.

218

THE .JOURNALO F CELL BIOLOGY • VOLUME 63, 1974

Downloaded from jcb.rupress.org on July 31, 2015

lnQ-

With the assumption that total exchangeable K (R~) is 45 nmol/106 cells, values of 1 - (Rt/R~) were plotted on a logarithmic scale against time. As shown in the inset in Fig. 1, a good fit to a straight line was obtained, indicating that 42K influx is a first-order process into a single compartment, and that the assumption of 45 nmol of exchangeable K / l 0 s cells is reasonable. With 346 #m 3 as the mean cell volume of human P M N s (Hempling, 1973), 45 nmol exchangeable K/106 cells gives an intracellular concentration of 130 retool/liter of cells exchangeable K. This value is essentially the same as published concentrations of K in human leukocytes determined by chemical analysis (Baron and Roberts, 1963:100 m M / l i t e r cell water; Baron and Ahmed, 1969:137 m M / l i t e r cell water; Lichtman and Weed, 1969:118 m M / k g cell water; Cividalli and Nathan, 1974:119 r a M / liter cell water). Therefore, the single exchangeable compartment of K represents essentially all of the intracellular K. The mean K influx determined on cells from 13 donors was 0.69 nmol/106 cells × min ( ± 0.095 SEM). Assuming a mean cell volume of 346 #,m s, unidirectional K influx can be expressed as 2.0 m m o l / l i t e r cells × min. Cividalli and Nathan (1974) reported unidirectional K influxes of 1.3 to 1.6 mmoles/liter cell water × rain. Their method

Published October 1, 1974

0.06 o o

c E 0.04 T a,

u. 0.02

_z

,/

[LY$1NE] raM-'

0.1~

-o | L

o.o~

...i

[LEUCINE] m i d "

FIGURE 2 Lineweaver-Burke plots of amino acid influxes in human PMNs. Cells were incubated in media with various amino acid concentrations, lysine (2a), or leucine (2b) for l0 rain. Results are given as reciprocal picomoles/108 cells × min vs. reciprocal millimoles/liter. Points represent single determinations.

A m i n o A c i d Influxes Double reciprocal plots of lysine and leucine influxes vs. external concentration are shown in Figs. 2 a and b. The correspondence between the experimental points and linearity suggests influx of both amino acids by saturable transport systems. These data enabled approximation of the kinetic constants of the fluxes, Vr~ax and apparent K,,. These values, along with mean fluxes at 0.1 m M external amino acid concentration, are given in Table II. All determinations involved 10-min exposures to the labeled amino acids. The influxes of [l~C]lysine and leucine had not reached saturation in 15 min, but the net rate of entry had declined. Thus after incubation for 10 min in the tracer, unidirectional influx would be underestimated from net tracer accumulation. Nevertheless, as will be shown below, fluxes based on incubation of 10

min give valid estimations of relative flux rates. The Vmax and apparent Km found by Ukena and Berlin (1972) for lysine influx in rabbit P M N s were 0.025 nmol/10 ~ cells x min and 0.25 mM, respectively. The Vmax was considerably lower than that found here. Preliminary data indicate that lysine and leucine influxes are by separate mechanisms: 2 m M leucine had no effect on lysine influx cells in 0.1 m M lysine. In contrast, Tsan and Berlin (1971) found a marked inhibition of lysine influx by leucine in rabbit alveolar macrophages. The method for measuring influxes involved washing the cells free of extracellular label by centrifugation and resuspension in ice-cold buffered saline. To obtain an indication of the extent to

DUNHAM, GOLDSTEIN, AND WEISSMANN Transport of K and Amino Acids in Leukocytes

219

Downloaded from jcb.rupress.org on July 31, 2015

involved washing the cells by filtration rather than centrifugation. A 10-min preexposure of 10 -5 M ouabain inhibited K influx by 52% (mean, 13 donors). The range' of the ouabain-sensitive component of K influx (38-69% of total influx) was considerably less than the range of absolute total K influx (0.30-1.25 nmol/10 e cells x min). Increasing the ouabain concentration to 10 -3 M gave no further inhibition of K influx. Exposure to 10 -8 M ouabain for 10 min caused 77% of the inhibition observed in 10 -5 or 10 -3 M ouabain. With the procedures employed, ouabain binding was irreversible: the K influxmeasured on cells preexposed to ouabain, but washed free of unbound ouabain, was the same as the flux measured in the presence of 10 -4 M ouabain. As in red blood cells, K inhibited ouabain binding. Preexposure to 5 x 10 -7 M ouabain for 10 min in 1 m M K resulted in complete inhibition of total ouabain-sensitive influx, whereas incubation with 5 x 10 -7 M ouabain in 5 m M K resulted in 66% inhibition of the ouabain-sensitive K influx. By analogy with red cells, the ouabain-sensitive K flux will be referred to here as the active or pump flux and the ouabain-insensitive flux as the passive or leak flux. It has not been established that the K gradient in P M N s is maintained by active transport, as indicated for red blood cells (Glynn, 1956; Dunham and Gunn, 1972) and other systems. The designations of pump and leak fluxes are used operationally, though ouabain may well act on P M N s in the same manner as on red cells.

Published October 1, 1974

which intracellular label is removed during washing, influxes of K and lysine were measured at 4°C and c o m p a r e d with fluxes measured at the same time at 37°C. Both total K influx and lysine influx were reduced by 90% at 4°C. Also, as expected, most of the K influx at 4°C was ouabain insensitive. Therefore, during washing in the cold, there is some loss of intracellular tracer, but only a small fraction of that which entered during the incubation at 37°C.

P o t a s s i u m Influxes and Phagocytosis Potassium influxes were compared in control cells and cells exposed to zymosan or latex partiTABLE II

Amino Acid Influxes in Human PMNs Based upon lO-min Exposures to [l"C]Amino Acid

nmol / l O~ cells × min

Lysine Leucine

0.021 (n = 6) 0.019 (n = 3)

Kinetic constants:~

Vmax

gm (apparent)

retool~ liter cells × min

nmol / l O~ cells × rnin

mmol/ liter

0.061

0.33

0.9

0.055

0.22

0.7

* External amino acid concentration. Taken from appropriate extrapolations of the data in Fig. 2.

control

r/f///////////////////////////A

cytocholasin B

///////////////////////A

zymosan zymoson and cyto. B

I//I//"/////2 ~'./,//.,'////J

Ia t e x

1

latex and

cyto. B

t.'///////////f] I

I L

IO0

[g',"/'///////////t////////////'1 510 % K LEAK

0

50 % K PUMP

=

IO0

FIGURE 3 K influxes in human PMNs after various treatments as percent of K influxes, both leak (ouabain-insensitive) or pump (ouabain-sensitive), in untreated or control cells. Aliquots of ceils were exposed to zymosan particles (10 particles/cell) or latex particles (100 particles/cell) for 30 min before measurement of the fluxes. Some of these aliquots were preexposed to cytochalasin B (5 #g/ml) for 15 min before addition of the particles; total exposure to cytochalasin B was 45 min. Total K influxes and % pump flux were close to the mean given in the text for all experiments, n = number of replicated experiments.

220

THE JOURNAL OF CELL BIOLOGY • VOLUME 63, 1974

Downloaded from jcb.rupress.org on July 31, 2015

Fluxes at 0.1 raM*

cles for 30 min before m e a s u r e m e n t of the flux. The results of several experiments are s u m m a r i z e d in Fig. 3. The ouabain-sensitive or p u m p flux and ouabain-insensitive or leak flux are given as the percents of these fluxes in control cells. In cells exposed to zymosan, pump influx was reduced by more than 60% c o m p a r e d to that in the control cells. The reduction in the p u m p was significant at the 0.016 level. The leak influx was reduced a b o u t 20% ( P = 0.03). Similar results were obtained in cells exposed to latex; the pump was inhibited by more than 50% and the K leak was reduced by 20%. The effects of exposure to zymosan w e r e observed only if fresh autologous serum was present in the reaction mixtures. If serum pretreated by heating to 56°C for 30 rain was employed, only a slight inhibition of K influx was observed. K influx was unaffected in cells exposed to zymosan-treated serum (1 m g / m l ) in the absence of particles. To determine if inhibition of K influx was due simply to m e m b r a n e internalization as a consequence of phagocytosis, cells were treated with cytochalasin B (5 tsg/ml) for 10 rain before exposure to particles. As shown in Fig. 3, cytochalasin alone may have reduced both ouabainsensitive and -insensitive K fluxes slightly ( P = 0.09 for both). In cells exposed to zymosan after cytochalasin B treatment, the K p u m p flux was again reduced, but to an extent not significantly different from that observed after exposure to zymosan alone (P = 0.375). In contrast, there was little or no inhibition of the K p u m p in cytochalasin B-treated cells exposed to latex particles.

Published October 1, 1974

Amino Acid Influxes and Phagocytosis

TABLE Ill

Purified PMNs

To exclude the possibility that the results observed in the experiments described above were due to the presence of m o n o n u c l e a r cells in the leukocyte suspensions, P M N s were purified with a Hypaque-Ficoll gradient. The cell suspensions obtained contained 98% P M N s . The effects of exposure of these cells to zymosan on K and lysine TABLE IV Amino Acid Influxes in Human PMNs: Effect of Time o f lncubation in [l"C]Amino Acid on Apparent Unidirectional Influx and on Apparent Percent Inhibition o f Influx by Preexposure to Zymosan Particles for 30 min.

K and Lysine Influxes, Measured Simultaneously, in Control Cells, and in Cells after Exposure to Zymosan or Latex for 30 rain

Influxes Lysine

K pump

K leak

% of control

Zymosan Zymosan 66 30 37 Cytochalasin B* 102 71 95 Cyto B + zymosan* 63 22 34 Total control influxes: lysine, 0.022 nmol/10 e x min; K, 1.17 nmol/106 x min. (K pump, 60%). Latex Latex 67 48 81 Cytochalasin B* 89 98 90 Cyto B + latex* 86 95 80 Total control influxes: lysine, 0.014 nmol/106 x rain; K, 0.37 nmol/106 x min (K pump, 66%). * PMNs were preincubated with cytochalasin B (5 mg/ ml) at 37°C for 15 min. Cytochalasin was also present during exposure to particles.

Lysine Leucine

Incubation time in tracer

Control influx

Inhibition after zymosan

rain

nmol/ lO6 cells x rain

%

0.5 10 0.5 10

0.091 0.022 0.089 0.018

14 24 20 34

TABLE V K and Lysine Influxes in a Purified Preparation of PMNs

Influxes Lysine

K pump

K leak

%of control Zymosan* 80 16 75 Total control influxes: lysine, 0.018 nmol/108 × min; K, 0.92 nmol/106 x min (K pump: 59%).

Differential counts; 98% PMNs, 2% eosinophils. * PMN's were exposed to zymosan for 30 rain.

DUNHAM, GOLDSTEIN, AND WEISSMANN Transport o f K and Amino Acids in Leukocytes

221

Downloaded from jcb.rupress.org on July 31, 2015

In Table III are the results of experiments in which lysine influx, from an external concentration of 0.1 m M , was measured along with K influx in cells exposed to particles. Lysine influx was reduced in cells that had ingested zymosan or latex, but far less than the reduction in the K pump. In a total of seven experiments, in which K and a m i n o acid influxes were measured simultaneously after exposure to particles, the inhibition of K influx was significantly greater than the inhibition of amino acid influx (P = 0.03). Table I11 also shows t h a t t r e a t m e n t with cytochalasin B had no effect on lysine influx and only slightly reduced K influx. In cells exposed to zymosan after t r e a t m e n t with cytochalasin B, both the lysine and K pump influxes were nearly the same as in cells exposed to zymosan without prior cytocbalasin treatment. Since the methods employed results in an underestimation of unidirectional amino acid influx, it is possible that the extent of inhibition after phagocytosis is also underestimated. Accordingly, undirectional influxes of a m i n o acids were determined in resting cells and in cells after exposure to zymosan, based upon incubation times in the tracers of 0.5 and 10 min. The results are shown in Table IV. The fluxes calculated from the 10-min incubation

are underestimated four- to fivefold relative to those based on the 0.5-min incubation; shorter incubation times might reveal an even greater underestimation. Nevertheless, the extent of inhibition of amino acid influx after exposure to zymosan was not reduced. The levels of inhibition based upon 10-min incubations were actually greater than those from 0.5-min incubations. The i m p o r t a n t point to be derived from these results is t h a t the difference in inhibition of K and a m i n o acid influxes after endocytosis c a n n o t be ascribed to an underestimation of the inhibition of a m i n o acid influx.

Published October 1, 1974

influxes are shown in Table V, along with the absolute total influxes in the control cells. Both the absolute fluxes and the inhibition observed in cells exposed to zymosan were generally the same as those observed with suspensions containing 85% P M N s , and, indeed, the K pump was inhibited much more than in less purified preparations.

Effect o f Colchicine

TABLE VI

Effects of Colchicine on Inhibition of K and Amino Acid Influxes by Particle Ingestion Influxes Leucine

K pump

K leak

%of control Zymosan Zymosan 75 20 96 Colchicine 104 92 97 Colchicine + 76 20 79 zymosan Total control influxes: leucine, 0.013 nmol/106 x min; K, 0.69 nmol/10 s × min (K pump: 54%). Latex Latex 76 36 82 Colchicine 96 75 109 Colchicine + latex 72 66 57 Total control influxes: leucine, 0.015 nmol/106 x min; K, 0.44 nmol/10 s x min (K pump: 69%). PMNs were preincubated with colchicine (5 x 10 -e M ) at 37°C for 1 h; Colchicine was also present during exposure to particles.

222

the K pump was than without. Both a slightly greater colchicine pretreat-

DISCUSSION The experiments cited in this report were designed to determine the relationship between two functions of the P M N plasma membrane: solute transport and phagocytosis. Using methods that have been utilized for the measurement of K and amino acid fluxes in other cell types, we were able to measure the transport of these solutes in human P M N s under resting conditions and after exposure to various phagocytic stimuli. The mean unidirectional K influx in resting P M N s measured in the present study, 2.0 m m o l / liter cells x min, is similar to the K influxes recently reported by Cividaili and Nathan (1974) (1.3 1.6 m m o l / l i t e r cell water x min; three determinations). No other determinations of K influx in P M N s have come to our attention. The similarity between our calculated value for exchangeable intracellular K and previously reported chemical determinations of K concentration in P M N s is a basis for confidence not only in the value for exchangeable K but in the values for unidirectional influx as well. There was a large range in absolute K influxes in differen t preparations of cells in the present work. This may have been due to actual differences in the cells from the various donors, since no such variability was found between duplicate determinations within an experiment. Furthermore, despite variability in absolute fluxes between experiments, the relative effects to which major significance is attached were consistent. There might be concern that the K fluxes are underestimated due to loss of ~2K from the cells during washing. Cividalli and Nathan (1974) presented evidence for K loss from human leukocytes in the cold. Though Berlin and Wood (1964) found that exposure to 4°C for 2 h had no significant effect on K concentration in leukocytes, the procedures employed here for the preparation of the leukocyte suspensions used in the experiments might also be expected to cause loss of K from the cells. Lichtman and Weed (1972) have shown conclusively that any such losses are restored by subsequent incubation for 1 h at less at 37°C in physiological concentrations of K. All of the

THE JOURNAL OF CELL BIOLOGY . VOLUME 63, 1974

Downloaded from jcb.rupress.org on July 31, 2015

K and amino acid influxes were measured in resting cells, cells exposed to zymosan or latex particles, and cells preincubated with 5 x 10 -~ M colchicine for 1 h before exposure to particles. The results are shown in Table VI. After ingestion of zymosan or latex the amino acid influx was inhibited slightly. The K pump was inhibited somewhat more than usual with both zymosan and latex, whereas the K leak was uncharacteristically not inhibited with zymosan. Colchicine treatment alone had little or no effect on amino acid or K influxes. The important conclusion to be drawn from these results is that colchicine did not cause an increase in inhibition of K pump influx or amino acid influx after particle ingestion. The percent inhibitions of the K pump and lysine influx were the same in cells exposed to zymosan in the presence or absence of colchicine.

With latex the inhibition of markedly less with colchicine zymosan and latex produced inhibition of the K leak after ment.

Published October 1, 1974

iM °M

CoeEVmr Ca

or

E=

R T I n -Ci - . - -aM F Co °M

(3)

where ~M and °M are the unidirectional influx and efflux, Co and C~ are the outside and inside concentrations, E is the electrical potential difference, and F, R, and T have their usual meanings. With the values just given and 37°C, E is predicted to be 68 inV. If E were determined entirely by the K concentration difference, it would be 87 mV calculated by the Nernst equation. Therefore E predicted by flux ratio analysis is less than the K equilibrium potential. If Na permeability were relatively high, and the Na gradient were a major determinant of E, then E might not be as high as 68 inV. Nevertheless, the membrane potential predicted by a passive K influx, 50% of total influx, is not unreasonable. Finally the ouabain-insensitive K influx might consist in part of a ouabain-insensitive active K influx analogous to the ouabain-insensitive Na efflux in red cells (Hoffman and Kregenow, 1966; Sachs, 1971). Cividalli and Nathan (1974) found no effect of ethacrynic acid On K concentration in human leukocytes; this drug inhibits the ouabaininsensitive Na pump in red cells. The results of our experiments provide new descriptions of the functional topography of the surface membrane of human PMNs. The topography appears to have two dimensions: (a) a distribution with respect to one another of transport sites for different solutes: and (b) a functional relation-

ship between transport sites, particle binding sites, and areas of the membrane involved in endocytosis. Evidence bearing on the first aspect of cell surface topography is provided by the large difference between inhibition of amino acid influx and inhibition of K influx after exposure to phagocytic stimuli. After ingestion of zymosan or latex particles or after surface binding of zymosan in cytochalasin-treated cells (see below), the K pump was inhibited 50-75%, while simultaneously measured lysine or leucine influxes were inhibited 30% or less. Assuming that the observed inhibition of transport was due primarily to internalization or inactivation of the surface membrane as a consequence of particle ingestion, these results indicate a topographical separation between the two kinds of transport sites. K transport sites appear to be more closely associated with areas of the membrane involved in the binding and phagocytosis of certain particles than the transport sites of amino acids. This functional topography need not reflect static spatial heterogeneity, but may indicate a mobilization of K sites toward (or amino acid sites away from) areas of the surface involved in particle contact and/or ingestion. The data summarized in Table IV demonstrate that the differences between levels of inhibition of K and amino acid influxes were not due to an underestimation of the inhibition of amino acid influxes, though the absolute unidirectional influx of amino acids was indeed underestimated. After ingestion of zymosan, ouabain-insensitive K influx was always reduced, though there was considerably more variability than in the reduction of ouabain-sensitive K influx. In some experiments, the K leak was reduced little or not at all after particle ingestion, even though the usual inhibition of the K pump was observed (e.g., Table VI). Therefore, little significance can be attached to effects of phagocytic stimuli on the K leak. The reduction usually observed may reflect a decrease of the total surface area of the cell. Conversely, there may be a spatial heterogeneity in areas of the membrane permitting passive K flux. It has been suggested that in red cells a component of passive cation fluxes, like active fluxes, takes place through specific, selective sites (Dunham and Hoffman, 1971). The inhibition of passive K flux in PMNs by membrane internalization is consistent with a mediation of the passive fluxes by such heterogeneously distributed specific sites.

DUNHAM, GOLDSTEIN, AND WEISSMANN Transport of K and Amino Acids in Leukocytes

223

Downloaded from jcb.rupress.org on July 31, 2015

experiments in the present study involved incubations at 37°C for 30-60 min before measurement of K influx. A large fraction of K influx, about one half, was insensitive to ouabain. This was not due to a low affinity for ouabain, since 10-5 M and 10-3 M ouabain gave the same levels of inhibition. If one makes the common assumption that the ouabainsensitive K influx is active and the ouabain-insensitive flux is passive, one can in simple thermodynamic terms approach the question: is it likely that there is a passive K influx of 1.0 mmol/liter × rain? If the cells are in steady-state with respect to K and all of K efflux is passive, K efflux will equal total K influx, or 2.0 mM/liter × min. From the fluxes and the K concentrations, 130 mM inside and 5 mM outside, one can calculate the electrical potential difference across the membrane with the flux ratio equation (Ussing, 1949):

Published October 1, 1974

224

treatment) both zymosan and latex are ingested irrespective of differences in surface binding, and the internalization of the surface membrane that accompanies ingestion results in a reduction of solute transport. In the nonphagocytic system (cytochalasin B-treated PMNs) surface binding of zymosan but not latex occurred, and this was sufficient to inhibit K and lysine influx. In both systems, inhibition of K influx exceeded inhibition of lysine influx. Thus there appears to be a spatial separation of sites responsible for lysine and K transport as well as an association of K pump sites with areas of the PMN surface membrane involved in particle ingestion, and specifically with zymosan-binding, perhaps C3b receptor sites. There may be a static association of K pump sites and C3b receptors in restricted areas of the membrane; or, conversely, the K pump sites may be mobilized toward the receptor sites when these latter sites are occupied. The inhibition of K transport by occupation of zymosan-binding sites may be steric: the zymosan particles may simply mask the K sites. It might also be allosteric, resulting from a conformational change in the K pump sites mediated through the zymosan-binding sites. In this instance, the transport sites may not be inhibited, but rather may have a reduced affinity for the substrate, thus reducing transport at subsaturating concentrations of substrate. A related possibility is that binding of particles to the surface may change the fluidity of the cell membrane, which may in turn affect transport sites. While the interpretation of the results in terms of membrane topography is consistent with the results, and appears to us the most likely context for explaining the results, another possibility should be considered. Phagocytic stimuli alter the metabolism of PMNs, most strikingly by stimulating CI oxidation of glucose (Skeel et al., 1969). Possibly the reduction of active transport accompanying particle ingestion is due to a redirection of metabolic energy away from transport mechanisms. Though this possibility cannot be ruled out, it must be considered in light of a large overall increase in metabolism in PMNs associated with phagocytosis (Sbarra and Karnovsky, 1959). There are two important differences between the present results and those obtained by Berlin's group on transport and phagocytosis. First, in the present study, transport was inhibited after phagocytosis of particles. Tsan and Berlin (1971) observed no inhibition after phagocytosis in otherwise untreated cells. Second, in the present system,

THE JOURNAL OF CELL BIOLOGY . VOLUME 63, 1974

Downloaded from jcb.rupress.org on July 31, 2015

Requirements for optimal phagocytosis of zymosan and latex particles differ in the systems employed in this study. Fresh serum must be present in the case of zymosan, since opsonization is a prerequisite for ingestion of this particle. There is considerable evidence that a portion of the third component of serum complement, C3b, is the factor that promotes adherence of these particles to the PMN surface membrane via specific membrane receptors (Winkelstein et al., 1972; Johnston et al., 1973; Lay and Nussenzweig, 1968). The lack of inhibition of K influx observed in PMNs exposed to zymosan in the presence of heated (complement-inactivated) serum is therefore expected. Fresh serum is not an absolute requirement for the ingestion of polystyrene latex particles, and, though opsonization by immunoglobulins has been suggested, specific binding of latex to PMNs has not been demonstrated (Van Oss and Stinson, 1970; Skeel et al., 1969). The second aspect of the functional topography of human PMNs indicated by our results is a complex relationship between transport sites, particle-binding sites, and areas of the membrane involved in phagocytosis. Support for this relationship comes from the results of experiments in which cytochalasin Btreated PMNs were exposed to zymosan or latex particles before measurment of the solute influx. Cytochalasin B, at the concentration used in these experiments, inhibits the ingestion of most particles by human PMNs (Davies et al., 1973; Zurier et al., 1973). Nevertheless, particles (particularly zymosan) can adhere to cells treated with this compound (Zurier et al., 1973). Thus the effects on solute transport of particle binding per se can be distinguished from effects of ingestion and membrane internalization. Cytochalasin B-treatment had little effect on lysine and K influx in resting cells. Zigmond and Hirsch (1972) similarly reported a lack of effect of cytochalasin B on leucine influx in horse leukocytes, though deoxyglucose transport was strongly inhibited. The inhibition of both lysine and K pump influx observed when untreated cells were exposed to zymosan was also seen with cytochalasin B-treated cells exposed to this particle. As in the former case, inhibition of K influx was again greater than inhibition of lysine influx. In contrast, when cytochalasin B-treated PMNs were exposed to polystyrene latex, no inhibition of solute influx was observed. Our interpretation of these results is as follows: In normal cells (without cytochalasin B

Published October 1, 1974

This work was supported by grants from the National

Institutes of Health (AM 11949 and AM 16196) and the Whitehall Foundation. Dr. Goldstein is the recipient of National Institutes of Health Special Fellowship (IF03 CA 55226-01) from the National Cancer Institute. Received for publication 15 November 1973, and in revised form 7 June 1974.

REFERENCES BARON, D. N., and S. A. AHMED. 1969. intracellular concentrations of water and of the principal electrolytes determined by analysis of isolated human leucocytes. Clin. Sci. (Oxfi ). 37:205. BARON, D. N., and P. M. ROBERTS. 1963. The sodium, potassium and water content of isolated normal human leucocytes. J. Physiol. (Lond.). 165:219. BERLIN, R. D., and W. B. WOOD. 1964. Studies on the pathogenesis of fever. XII. Electrolytic factors influencing the release of endogenous pyrogen from polymorphonuclear leucocytes. J. Exp. Med. 119:697. BERRY, R. W. and M. L. SHELANSKt. 1972. Interactions of tubulin with vinblastine and guanosine triphosphate. J. Mol. Biol. 71:71. BOYUM, A. 1967. Isolation of mononuclear ceils and granulocytes from human blood. Scand. J. Clin. Lab. Invest. 97(Suppl.):77. BRAY, G. A. 1960. A simple efficient liquid scintillator for counting aqueous solutions in a liquid scintillation counter. Anal. Biochem. 1:279. CIVIDALLI, G., and D. G. NATHAN. 1974. Sodium and potassium concentration and transmembrane fluxes in leukocytes. Blood J. Hematol. 43:861. DAVIES, R., R. 1. Fox, M. POLYZONIS,A. C. ALLISON, and A. D. HASWELL, 1973. The inhibition of phagocytosis and facilitation of exocytosis in rabbit polymorphonuclear leukocytes by cytochalasin B. Lab. Invest. 28:16. DUNHAM, P. B., and R. B. GUNN. 1972. Adenosine triphosphatase and active cation transport in red cell membranes. Arch. Intern. Med. 129:241. DUNHAM,P. B., and J. F. HOFFMAN.1971. Active cation transport and ouabain binding in high potassium and low potassium red blood ceils of sheep. J. Gen. Physiol. 58:94. GLYNN, I. M. 1956. Sodium and potassium movements in human red cells. J. Physiol. (Lond.). 134:278. HEMPLING, H. G. 1973. Heats of activation for the exosmotic flow of water across the membrane of leucocytes and leukemic cells. J. Cell. Physiol. 81:1. HENSON, P. M. 1971. Interaction of cells with immune complexes: adherence, release of constituents, and tissue injury. J. Exp. Med. 134(3, Pt. 2):114s. HOFFMAN, J. F., and F. M. KREGENOW. 1966. The characterization of new energy dependent cation transport processes in red blood cells. Ann. N. Y. Acad. Sci. 137:566. JOHNSTON, R. B., JR., S. L. NEWMAN, and A. G.

DUNHAM, GOLDSTEIN, AND WEISSMANN Transport of K and Amino Acids in Leukocytes

225

Downloaded from jcb.rupress.org on July 31, 2015

disruption of microtubules had no effect on the inhibition of transport association with particle ingestion, whereas Ukena and Berlin (1972) found that after colchicine treatment, phagocytosis did result in inhibition of transport. These differences may be attributable to one or more factors. First, cells from different species of m a m m a l were used: human cells in the present study and cells from rabbits in the studies of Berlin's group. Second, in the present study, P M N s were obtained from peripheral blood. Berlin's group used alveolar macrophages and peritoneal PMNs. The experiments involving treatment with colchicine were performed only with PMNs. Finally, in the present work, the cells were in suspension, whereas Berlin's group used cells that had been allowed to adhere to glass. Attachment to glass may be associated with changes in the functional state of PMNs. For example, it has been suggested that microtubules are involved in a t t a c h m e n t and spreading of m a c r o p h a g e s (Reaven and Axline, 1973). Attached cells might therefore respond to colchicine differently from cells in suspension. Recently Berlin's group reported further evidence on regulation of membrane topography that is in part complementary to that given here (Oliver et al., 1974). It was found that lectin-binding sites in rabbit P M N s were internalized selectively during phagocytosis of droplets of an oil emulsion. The evidence for selective internalization was a reduction of lectin-binding sites per milligram of membrane protein after phagocytosis. Thus, these results correspond to the selective internalization of K transport sites reported here. Pretreatment with colchicine for 30 rain at 5 × 10 6 M did not inhibit phagocytosis, but did eliminate the selective internalization of lectin-binding sites. In contrast, colchicine did not have a corresponding effect on the selective internalization of K transport sites in the present study. It is important to point out that this last study from Berlin's group used P M N s in suspension rather than attached to glass, as in their previous work. The interpretation of the results on lectin binding may be complicated by the fact that at 37°C, lectins induce their own internalization by a process that remains unclear (Oliver et al., 1974), whereas P M N s do not otherwise appear to internalize membrane in the absence of specific phagocytic stimuli ~Henson, 1971 ).

Published October 1, 1974

226

tosis on membrane transport of nonelectrolytes. J. Exp. Med. 134:1016. UgENA, T. E., and R. D. BERLIN. 1972. Effect of colchicine and vinblastine on the topographical separation of membrane functions. J. Exp. Med. 136:1. USSlNG, H. H. 1949. The distinction by means of tracers between active transport and diffusion. Acta Physiol. Scand. 19:43. VAN OSS, C. J., and M. W. STINSON. 1970. lmmunoglobulins as specific opsonins. I. The influence of polyclonal and monoclonal immunoglobulins on the in vitro phaocytosis of latex particles and staphylococci by human neutrophilis. J. Reticuloendothel. Soc. 8:397. WEISENBERG, R. C., G. G. BORISY, and E. W. TAYLOR. 1968. The colchicine-binding protein of mammalian brain and its relation to microtubules. Biochemistry. 7:4466. WEISSMANN, G., R. B. ZURIER, P. J. SPIELER, and I. M. GOLDSTEIN. 1971. Mechanisms of lysosomal enzyme release from leukocytes exposed to immune complexes and other particles. J. Exp. Med. 134(3, Pt. 2):149 s. WINKELSTEIN,J. A., H. S. SHIN,and W. B. WOOD. 1972. Heat labile opsonins to pneumococcus. Ili. The participation of immunoglobulin and of the alternate pathway of C3 activation. J. Immunol. 108:1681. ZIGMOND,S. H., and J. G. HIRSCH. 1972. Cytochalasin B: inhibition of D-2-deoxyglucose transport into leukocytes and fibroblasts. Science ( Wash. D.C.). 176:1432. ZURIER, R. B., S. HOEESTEIN, and G. WEISSMANN. 1973. Cytochalasin B: effect on lysosomal enzyme release from human leukocytes. Proc. Natl. Acad. Sci. U.S.A. 70:844.

THE JOURNAL OF CELL BIOLOGY - VOLUME63, 1974

Downloaded from jcb.rupress.org on July 31, 2015

STRUTH. 1973. An abnormality of the alternate pathway of complement activation in sickle-cell disease. N. Engl. J. Med. 288:803. LAY, W. H., ancl V. NUSSE.NSWEIG. 1968. Receptors for complement on leucocytes. J. Exp. Med. 128:991. LICHTMAN, M. A., and R. I. WEED. 1969. The monovalent cation content and adenosine triphosphatase activity of human normal and leukemic granulocytes and lymphocytes: relationship to cell volume and morphologic age. BloodJ. Hematol. 34:645. OLIVER, J. M., T. E. UKENA, and R. D. BERLIN. 1974. Effects of phagocytosis and colchicine on the distribution of lectin-binding sites on cell surfaces. Proc. Natl. Acad. Sci. U.S.A. 71:394. REAVEN, E. P., and S. G. AXLINE. 1973. Subplasmalemreal microfilaments and microtubules in resting and phagocytizing cultivated macrophages. J. Cell Biol. 59:12. SACHS, J. R. 1971. Ouabain-insensitive sodium movements in the human red blood cell. J. Gen. Physiol. 57:259. SBARRA, A. J., and M. L. KARNOVSKY. 1959. The biochemical basis of phagocytosis. I. Metabolic changes during the ingestion of particles by polymorphonuclear leukocytes. J. Biol. Chem. 234:1355. SIEGEL, S. 1956. Nonparametric Statistics. McGrawHill, Inc., New York. 312. SKEEL, R. T., R. A. YANKEE, W. A. SPIVAK, L. NovlKovs, and E. S. HENDERSON. 1969. Leukocyte preservation. 1. Phagocytic stimulation of the hexose monophosphate shunt as a measure of cell viability. J. Lab. Clin. Med. 73:327. TSAN, M.-F., and R. D. BERLIN. 1971. Effect ofphagocy-